PB1133 - Validating the JAKPOT prediction rule for identifying which patients with erythrocytosis are unlikely to have a JAK2 mutation

Tuesday, June 25, 2024

13:45 – 14:45 ICT

Room: Exhibition Hall

Presenting Author(s)

Barbara D. Lam, MD

Fellow
Beth Israel Deaconess Medical Center
Boston, Massachusetts, United States

Co-Author(s)

KO

Kevin O'Sullivan

Resident
Beth Israel Deaconess Medical Center
boston, Massachusetts, United States
RL

Roosevelt Lu

Research Assistant
Beth Israel Deaconess Medical Center
boston, Massachusetts, United States
JF

Jason A. Freed

Assistant Professor
Beth Israel Deaconess Medical Center
Boston, Massachusetts, United States
GG

George Goshua

Assistant Professor
Yale Cancer Center
New Haven, Connecticut, United States
JR

Justine Ryu

Assistant Professor
Yale Cancer Center
New Haven, Connecticut, United States
Rushad Patell, MBBS

Assistant Professor
Beth Israel Deaconess Medical Center
Boston, Massachusetts, United States

Background: Erythrocytosis has many causes, including Polycythemia Vera (PV). PV, a type of myeloproliferative neoplasm (MPN), is often driven by a JAK2 mutation and associated with an increased risk of thrombosis. The prediction rule JAKPOT was developed in 2022 to determine which patients with erythrocytosis benefit from JAK2 testing. Patients who met zero criteria (red blood cell count >6.45x1012/L, platelets >350x109/L, and neutrophils >6.2x109/L) were JAKPOT ‘low-risk’ for harboring a JAK2 mutation. JAKPOT has not yet been externally validated.

Aims: Externally validate the JAKPOT rule on a cohort of 1000 patients at a tertiary care center in Boston, Massachusetts, USA.

Methods: Demographics and laboratory data for 1000 random patients with JAK2 testing were extracted from the hospital data warehouse. Manual chart review identified patients who met inclusion criteria (those evaluated for erythrocytosis who had same-day JAKPOT variables available) and those who met exclusion criteria (a known diagnosis of MPN). Negative predictive value (NPV) was calculated using Python version 3.9.

Results: Of 1000 patients, 723 were excluded (512 referred for reasons other than erythrocytosis, 153 missing same-day JAKPOT variables, 29 cancelled tests, 29 known MPN diagnoses). Validation of JAKPOT on the remaining 277 patients demonstrated a 98.4% NPV (95% CI: 0.95-0.99). Of the 34 patients found to have a JAK2 mutation, 31 met at least one JAKPOT criteria and all were diagnosed with PV or a PV/Essential Thrombocythemia overlap syndrome.

Conclusion(s): The JAKPOT rule is effective at identifying patients with erythrocytosis who are at low risk of harboring a JAK2 mutation. Next steps include expanding the validation cohort to other institutions and conducting a cost-effectiveness analysis.